Controlled studies of surgical treatment of men with angina pectoris due to substantial coronary vascular disease have shown significant differences in subsequent survival rates in only a minority of the patients. Clinical and non-invasive exercise predictors of unoperated men with symptomatic coronary heart disease have identified a major gradient of risk of sudden cardiac death with as few as three variables representing size and functional limits of the left ventricle. These variables may represent better criteria for selection of patients for aortocoronary bypass surgical treatment than the conventional sizes of myocardial ischemia and/or vascular disease. This study proposes to measure both the limits of cardiac output (Q max) and the peripheral extraction of oxygen (D AVO2) at symptom-limited maximal exercise preoperatively and again between 7 and 12 months postoperatively to determine the changes in cardiac capacity and the possible relationship to subsequent differences in CHD mortality over 3-5 years in relation to the classification of preoperative arteriographic as well as resting and exercise ventricular variables.